Research On The Consistency Maintenance Of CVEs

Consistency maintenance is a kernel problem in Collaborative Virtual Environments (CVEs) research. Dead-Reckoning (DR) algorithm and Local-lag mechanism are two methods to realize consistency control in Virtual Environments (VEs). DR algorithm is most widely used. However, DR algorithm will cause many short-term inconsistencies, so it doesen't suit, for highly interactive tasks. Local-lag mechanism can effectively eliminate short-term inconsistencies, and can meet the demand of highly interactive task. Therefore, it is a promising algorithm. But there are still many problemes need to resolve about Local-lag algorithm. This paper tries to improve local-lag algorithm by studying it through experiments. The contents of this paper are as follows:In order to study the consistency maintenance in CVE, we design an desktop collaborative virtual envoirnment, named JBCVE. JBCVE has very short reponse time and can surport highly interactive task. In addition, it can be quickly adjusted according to the requirements of experiment. It is a good experiment platform.Choosing a proper lag value is a key problem in local-lag mechanism. This paper studies the effects of lag value (0 ms ~ 900 ms) on task performance in a desktop CVE system. Experiment results indicate that the effect of lag value on task performance is not linear. The effect could be separated into four segments by three dividing points: 150 ms, 300 ms and 600 ms. Lag has no effect on task performance while ranging from 0 ms to 150 ms. From 150 ms to 300 ms, lag slightly affects task performance. Lag deteriorates task performance seriously while ranging from 300 ms to 600 ms. When lag is larger than 600 ms, the task can't be accomplished sometimes. Local-lag mechanism sacrifices some responsiveness of the system and leads to the deterioration of task performance, especially when lag is large. This paper proposes an echo method to improve the task performance. The echo method shows some system information to subjects to help subjects to control objects. Experiment results show that the echo method does improve the task performance. The resultsalso give support to the guideline that properly showing some system information will increase the usability of the system.The local-lag algorithm has not the ability of filtering data. This paper proposes a new data filtering method by analyzing the network transmitting messages in CVE. The method categorized the network transmitting messages into key messages and non-key messages. For non-key messages, method filters them adaptively according to network condition. For key messages, method guarantee them be transmitted reliably.